华南理工大学学报(自然科学版) ›› 2009, Vol. 37 ›› Issue (2): 70-76.

• 交通运输工程 • 上一篇    下一篇

多车型、多准则、混合策略的网络均衡行为

刘伟铭 付凌峰 刘驰   

  1. 华南理工大学 土木与交通学院, 广东 广州 510640
  • 收稿日期:2008-01-16 修回日期:2008-03-13 出版日期:2009-02-25 发布日期:2009-02-25
  • 通信作者: 刘伟铭(1963-),男,教授,博士生导师,主要从事复杂交通系统的优化、处理、识别与控制研究. E-mail:mingweiliu@126.com
  • 作者简介:刘伟铭(1963-),男,教授,博士生导师,主要从事复杂交通系统的优化、处理、识别与控制研究.
  • 基金资助:

    国家西部交通建设科技项目(200139822382)

Network Equilibrium Behaviors with Multiple Vehicle Classes, Criteria and Mixed Strategies

Liu Wei-ming  Fu Ling-feng  Liu Chi   

  1. School of Civil Engineering and Transportation, South China University of Technology, Guangzhou 510640, Guangdong, China
  • Received:2008-01-16 Revised:2008-03-13 Online:2009-02-25 Published:2009-02-25
  • Contact: 刘伟铭(1963-),男,教授,博士生导师,主要从事复杂交通系统的优化、处理、识别与控制研究. E-mail:mingweiliu@126.com
  • About author:刘伟铭(1963-),男,教授,博士生导师,主要从事复杂交通系统的优化、处理、识别与控制研究.
  • Supported by:

    国家西部交通建设科技项目(200139822382)

摘要: 建立了一个多车型、多准则、混合策略网络均衡行为的变分不等式模型,来预测道路收费和交通信息系统作用下公路交通网络中的流量分布形态.该模型在依据车型和策略对出行用户进行划分的基础上,通过行程费用参数的设置反映不同车型用户在行驶速度、行驶费用、道路收费等方面的差异,通过设置感知行程费用函数反映不同策略出行用户在进行路径选择时优化目标的差异,并提出这种情况下的网络均衡条件.文中还介绍了采用改进连续平均算法进行模型求解的步骤.算例分析表明,该模型可以描述道路收费和交通信息系统对于提高网络效率所发挥的作用,并揭示混合策略网络均衡行为的一些特性.

关键词: 交通流量, 混合均衡, 多类型用户, 多准则, 变分不等式

Abstract:

In this paper, a variational inequality (VI) model is proposed to describe the network equilibrium be- haviors with multiple vehicle classes, multi-criteria and mixed strategies. The proposed model is then used to predict the flow distribution in a highway transport network under the condition of road tolling and traffic information system. In the proposed model, the users are classified according to the vehicle type and the route choice strategy that every individual user employs; the travel cost function parameters are set to reflect the differences in users with distinct vehicle types in terms of travel speed, travel cost and road toll, etc. ; the perceived travel cost functions are set to describe the variations of optimization object in route choice for users with different travel strategies; and the equilibrium condition for this specific situation is also determined. Moreover, the solution procedures for the VI model by means of an improved successive average algorithm are introduced. Numerical experiment results indicate that the proposed model well describes the roles of road tolling and traffic information system played in the network efficiency improvement, and further reveals some characteristics of the network equilibrium behaviors with mixed strategies.

Key words: traffic flow, mixed equilibrium, multi-type user, multi-criterion, variational inequality